buying an oscilloscope

[MitchK]

Hey folks I'm looking at buying an oscilloscope, there have been several occasions now where it seems it could have really helped me out so I think its time.

I really don't know much about them though so I'm wondering if anyone has an recommendations for ones to look for or stay away from, and what I'll need to troubleshoot guitar amps.

one thing I really don't get is that some are 100MHz some are 40MHz and so on, is this just the upper limit as to what it can register, is this of any concern when dealing with audio frequency?

Also, am I going to need to buy a signal generator as well?

[Enzo]

In my opinion, you don't NEED a signal genny, but they are very handy. That said, you can MAKE a simple one with a chip or two and a 9V battery. For basic repairs, mostly any signal will do. Having a sine wave generator where you can adjust the level and freq to anything in the audio band will be more of a necessity when you get into characterizing performance. My main bench genny sits on 100Hz almost all the time. I use music more than I use the genny. With music I can hear a hole in the response curve faster than I can find it with sweeps. And of course I use the guitar.

I am not suggesting signal generators are bad or useless, I do use mine a lot. Crossover distortion settings would be hard without one. But the scope will widen your horizons substantially without one. With one would be better, yes. The purchase of a scope is not wasted without one is what I am trying to say. Any more than your voltmeter is wasted without a scope. Plenty of good work was done with just that meter, right?

A scope is like a visual voltmeter. You can watch the voltage in a circuit vary continuously. It lets you see gross distortion. A missing half a waveform is pretty obvious on a scope, but unless your ear is experienced to recognize that particular form of distortion, not as easy without it. A symptom like that instantly tells you one side of a p-p amp is not running. A scope can help see frequency response, but since guitar amps are pretty well out of it over 5kHz, this is not so helpful, to me anyway. CLipping is easy to see, even clipping you can't hear. Crossover distortion becomes obvious.

SInce the bandwidth of guitar amps is so limited, there is no need for a 100MHz scope. Certainly one will work just fine, but so will a 10MHz scope or even a 500kHz scope. Oh, someone can come up with some arcane procedure that will only work with a whiz bang scope, but by the time you REALLY need something like that, then you will be ready for a scope upgrade anyway.

Unless it is some fine old piece of history, I doubt you will even find a scope under 10MHz. Don't worry about the MHz, you will have enough no matter what. Get as much as you like, just don't pay a lot extra for more scope than you need. A working 20MHz scope next to a working 100MHz scope - if they are priced the same, sure I'd take the 100, but when the one is $100 and the other is $300, then I'd take the slow one all day.

Yes, the scope freq rating is the max it will reliably respond to, it will always work at lower speeds. Other than DC, audio is about as slow as it gets. SOme applications require higher speed, like computer circuits and other high speed digital stuff. Or microwave and UHF radio.

ALmost all scopes these days are triggered sweep. All that means is the trace on the screen will synchronize with the signal so it appears stable. Like a strobe light makes a spinning fan appear stable if it blinks at the same speed as the fan turns. if a scope triggers the trace to sweep across the screen at the same point of each waveform, then the picture will appear in the same position. Some older scopes had "recurrent" sweep, which mean the trace just swept at whatever speed you set the knob to. A little tweaking and the waveform can be fairly steady, so even those old dinosaurs will still serve you well. These might be the scopes you can find for $20 at a ham fest. Working even. Triggering is preferred, but not essential.

Vertical sensitivity. A scope is really just an amplifier - or maybe a preamp - and instead of pushing a speaker back and forth, it pushes a beam of electrons back and forth on the scope screen. That is the glowing line. The vertical knob is just a volume knob. The moer you turn it up, the more sensitive it becomes. Some scopes are VERY sensitive. All the way up might need only 1 millivolt or even less to move the screen trace a whole division on the screen. Turn your scope up that far, and you get nothing useful, just the same noise and crap a guitar cord picks up without a guitar. Older scopes often went up as high as 10 or more volts per screen division. Many newer scopes only go up to 5 volts/div. I prefer the larger voltages. Let me detour for a moment

By the way division merely means the little cross lines on the screen. They are usually about a centimeter apart. They divide the screen. 5v/div means if your probe encounters 5 volts, it will move the screen trace one division - one centimeter - on the screen. My scope here is typical at 8 divisions tall and 10 wide.

Probes. Scopes, especially used ones, are often sold without probes. You will need at least one. You want a 10x (10 times) probe, or better yet, a switchable 10x/1x probe. You can spend $1000 on a probe, but the common cheap $30 probes have always worked fine for me.

A 10x probe just has some high value resistors in it as a voltage divider. What ever voltage the tip of the probe encounters, it will be divided by 10, so the scope itself sees 1/10 as much. So a 100v test point will seem like 10v to the scope. This way I can probe 550V and teh scope will see 55v. As long as the probe is rated for a higher voltage, a lower voltage scope can view it.

When I have a high voltage ar a large signal, I would like to have it fit on my scope screen. If the lowest sensitivity on the scope is 5v/div, then on my 8 division screen, I am limited to 40 volts. I can move the trace around, so centered i can then see 40v peak to peak on a signal, or set the trace to the bottom, and I can see 40 volts of DC. Even on my 10x probe, that still limits me to 400v. If my scope goes down to 10v/div, then I have no problem fitting 600v on my screen. Usually there are other ways around this, but you get the idea. SO I prefer a scope with a 10k/Div scale.

By the way, when you use the 10x probe, you set the scope to 5v/div, but assume 50v/div while looking at the trace.

Most modern scopes are dual trace, meaning you can watch two different inputs, either switching between them, or superposing them together on the screen. It can be handy sometimes, but mostly I use just one channel. I do keep my second channel connected to my bench speaker/load panel. DOn't turn down a single channel scope if it otherwise suits just for that one quality. And that 100MHz scope that can be a four channel? I am sure you could come up with some application to use all four traces, but really, no one needs that to service guitar amps.

My old old B&K has been a workhorse for me, certainly Tektronix scopes are first rate, but inexpensive Hitachi scopes worked well for us, and others report success even with things like the MCM house brand Tenma.

[drewl]

Almost all scopes have a "cal" output, or a square wave test point on the front panel for checking the scope's accuracy.
You can use this instead of a signal generator for troubleshooting, but it is a SQUARE wave, so you wouldn't be able to tell if a certain stage was clean (no distortion) or not.

[bob p]

some helpful threads that contain good reference info:

Time to get an O-scope...tips?

Oscilloscope Roundup

Signal Generator Advice Needed

Lissajous Figures - Indulge Your Inner Geek

hth

Enzo, would you mind if I quoted your excellent summary and added it to the "Time to Get an O-Scope" thread?

[Enzo]

Well, it was a VERY informal discussion. Feel free to quote anything I post, no secrets here.

[Mars Amp Repair]

As for 100mhz scopes, I agree about 20MHz being more reasonable.

Reasons...I repaired camcorders & all kinds of microprocessor controlled devices for years & mostly had to keep the 20mhz scope frequency limiter on...thats a switch that limits the response to 20mhz so you don't have to determine what is the actual signal & what is just a bunch of very high frequency noise & crap that the device isn't going to respond to anyway.

BTW I have a Tektronix 465B 100mhz dual trace delayed sweep wiz bang scope that believe it or not was given to me. I've never taken the freq limiter off of 20MHZ for amps or anything in the music biz. glen

[Enzo]

I wonder if the frequency limiter is just a fancy name for a bright switch.

The only time I ever wished my scope had more bandwidth was working on a wireless unit. belt pack sending at like 179MHz. I found if I wrapped a couple feet of wire around the transmitter and connected my scope probe to it, I could just get a tiny little trace deflection on my 35MHz scope. I couldn't resolve any waveform, but it was enough of an indication that the transmitter was sending out at least a carrier. I found my old Radio Shack all band scanner picked up a lot of the wireless units.

[mlannoo]

Ted's got a good price on a general all purpose oscope - buying new is nice because you don't have to worry if it's been abused or needs calibration.
https://amptechtools.powweb.com/scope.htm

I've always found Tone-Lizard's site informative and amusing - Here's a pleasant read about oscope's:
http://www.tone-lizard.com/Oscilloscopes.htm

[Mars Amp Repair]

This is an Instek scope. Tenma has the identical scope for about $100 more. I have a triple dc power supply by Instek & it's identical to the Tenma version...
Still the AmpTech price is lower than the Tenma or Instek by almost $100.

It's obvious that these are all made by one manufacturer & many others put their name on them. Not certain who's the originator, tho. Either way, I've used a few of the Tenma/Instek scopes. For the price, they perform quite nicely. g

[Mars Amp Repair]

Scope brite switch....why not....
only problem is I don't have the wizz bang Tektronix 3 billion dollar 100MHZ scope probes...don't need them, so the probes I guess act like a built-in brite switch. Mine cost $15/pr off e-bay...just wish at times that they had a x1 setting. They're just x10 probes. I've got one older probe that has x1 & x10 for special occassions. g

[MitchK]

excellent! thanks very much, but just to make things clear on the v/div

so for example I'm seeing a lot of ones that are 5v/div and have 10 divisions, 5 on either side with a 10x probe that would really mean the most I can look at is 250v? which isn't going to be very helpful in working on amps that are running at 500V+?

[nico.verduin]

That is why they use the divisor switch 1:10 which divides the input voltage by 10. Now allowing you to measure up to 2.5kV if your probe itself can handle it. Here in Holland there are a lot of Philips P3230 that can handle up to 600V direct on the inputs. They sell here for about 15uS$ because they are big, heavy and not sexy. They last about 3-4 years with intensive use. And then I'll just get a new old one.

[Enzo]

OK. You are assuming the trace at rest has to stripe across the center. Wanna look at 500VDC? Move the trace down to the bottom line Now all ten divisions are above you.

In fairness, I rarely look at all 500 volts of the DC. If I just want to see if it is present or not, I probably already did that with my voltmeter. I am more likely to look at the 500VDC to see if it is clean, looking for ripple. In that case, I set the input coupling switch to AC instead of DC. Now ONLY the ripple will deflect the trace. I can adjust the vertical so the 20 volts of ripple will fit in the screen. If you look at the entire 500v, then any ripple will be just tiny wiggles at the top.

And if you want to look at something taller than the screen, nothing says you can dial the bottom down past the screen bottom. And in cases like looking at crossover distortion in a power amp, we will set the scope up so the sine wave output is WAAAYYY larger than the screen, since all we care about is the little crossover notch at the center. We want to see that little glitch large and visible, who cares about the rest of the waveform. You don't have to take your shirt off to trim your fingernails. Or some such old home metaphor.

My scope is usually on AC anyway. My voltmeter was the first line of defense. Got B+? check. Got cathode volts? check. Now I am looking for signal with the scope. I don't need to see the DC offset at the plate, just the signal riding on it. I COULD set it up so the entire 150v of B voltage at the plate is on screen with the signal on it. But a second later at the next stage frid, I don't gave that 150v, I only have signal, so I'd have to reconfigure the scope. I just follow the signals on AC and there I am.

Now if you have some voltage that slides down or something, then sure, look at the whole thing. Obviously the application makes a difference.

ANother thing about running in AC mode. I can look for ripple on the B+, then flip over and read the ripple on the bias supply. If I were in DC, I'd have to stop and move the trace to the top so the negative voltage would send it DOWN across the screen. Otherwise the bias would be off screen below.